Published September 2015.
As part of the Smart Data Solution Center Baden-Württemberg (SDSC-BW), the Karlsruhe Institute of Technology (KIT) conducted a potential analysis for leather-offcut optimization on behalf of Rolf Benz in the Smart Data Innovation Lab.
Within the process of evaluating and planning leather cuttings, diverse data become available. The priority will be the quality of the used animal hides already fine-grained recorded by the company. The fully automated cutting machines and resource planning systems (ERP) provide additional data in the form of order and production data that are used in different contexts of the operative procedure and interrelated in various ways. Beside the evaluation and cutting logs of the processed animal hides from an eight-month period, the potential analysis (see below) of the SDSC-BW examined the operating data of performed productions and upholsteries.
The cutting optimization of natural leather is a highly sophisticated process that already factors in the different quality characteristics of the leather hides while placing the cutting templates. Not only the underlying natural product leather but also the furniture manufactured from it – individually and just-in-time – are available in many variants. For this reason, a flexible and individual consultation regarding Smart Data was especially important for Rolf Benz as an individual manufacturer. Over the years, the company has developed various strategies to improve the conditions for the automated placement of the cutting templates – among others, by bundling orders and pre-sorting the leather. Because: Every percent less offcut is a relevant competitive advantage. By means of machine learning methods, the SDSC-BW worked on finding approaches for automatizing process improvements previously performed manually.
Within the scope of the potential analysis conducted free of charge, the experts of the SDSC-BW studied all provided production data over an eight-month period. The initial analysis focused on classifying external/systematic factors or conforming already known factors in order to identify the potential for a possible offcut reduction. The analysis was funded by the Ministry of Science, Research, and Arts Baden-Württemberg.
Based on Smart Data-methods, the expert team developed an approach that enables the automated generating and testing of hypotheses with consideration of further data. This Smart Data approach can be combined with a cutting simulation. Due to the close coupling of Smart Data analytics, optimization, and simulation, Rolf Benz hopes for a prompt implementation of this potential in the production process to further strengthen its site.
In addition to the results presented above, the four-week analysis of the SDSC-BW at Rolf Benz has shown how interesting the potential use of tools, such as process mining on operating data, is. Without further intervention into the data pool, but on the basis of production data acquisition, it allows Rolf Benz to purposefully visualize and detect potential bottlenecks in the downstream upholstery. The SDSC-BW presented the results of the potential analysis to the production experts of Rolf Benz and placed them at their disposal. Since the first project was successful, it is planned to continue the cooperation with the aim to gather higher resolution data and include simulations as an alternative or supplement to the previous historical data analysis.
More information can be found at: www.sdsc-bw.de
KIT, Rolf Benz, Sicos BW
Andreas Meier, email@example.com
Sept.– Nov. 2015